Glen Perkins likes math

[Brock Beauchamp]

It's just a shame that most of his teammates don't share his enthusiasm for the subject. This is a pretty decent primer on some of the more popular advanced metrics and how Perkins uses them to his advantage.

 

http://mlb.mlb.com/news/article/min/for-twins-closer-glen-perkins-baseball-really-is-a-numbers-game?ymd=20140804&content_id=87928284&vkey=news_min

[Craig Arko]

Smart Minnesota kid.

[drjim]

At a fundamental level, why does a player need to know this information? I understand why it is important to analyze for future performance/extensions/trades/etc, but I don't see what difference it makes when you are standing on the mound.

 

The main possible use would be if a guy was underperforming his peripherals and it could be cited to boost his confidence but I suspect this is not what you have in mind when posting this.

[JB_Iowa]

While I'm impressed with Perkins, I'm also impressed that Bollinger wrote this article.

 

And while others on the team may not take the stats as much to heart as Perkins, it's encouraging that some of his teammates look to him for explanations.

 

I can only hope that Jack Goin is force feeding some of this to the coaching staff.

[Craig Arko]

At a fundamental level, why does a player need to know this information? I understand why it is important to analyze for future performance/extensions/trades/etc, but I don't see what difference it makes when you are standing on the mound.

 

The main possible use would be if a guy was underperforming his peripherals and it could be cited to boost his confidence but I suspect this is not what you have in mind when posting this.

“Victorious warriors win first and then go to war, while defeated warriors go to war first and then seek to win."

- Sun Tzu

[drjim]

“Victorious warriors win first and then go to war, while defeated warriors go to war first and then seek to win."

- Sun Tzu

 

Doesn't really answer the question. I fully expect Twins starters to undergo full preparation with a scouting report and plan of attack for every hitter every game.

 

I would hope that the scouting report includes a detailed breakdown of the results each hitter has against each type of pitch in all zones and counts. But I still don't see what value knowing the numbers cited in the article have for a pitcher as he prepares to pitch.

[old nurse]

Doesn't really answer the question. I fully expect Twins starters to undergo full preparation with a scouting report and plan of attack for every hitter every game.

 

I would hope that the scouting report includes a detailed breakdown of the results each hitter has against each type of pitch in all zones and counts. But I still don't see what value knowing the numbers cited in the article have for a pitcher as he prepares to pitch.

It is not for preparation to pitch, it is daily assessment of how he is pitching to stay on top of performance

[Mike Sixel]

Process control, like any other effort. He is doing statistical process control to understand the outcomes he is producing, to understand if he needs to change what/how he is doing it. Every successful business does this......

[Otto von Ballpark]

At a fundamental level, why does a player need to know this information? I understand why it is important to analyze for future performance/extensions/trades/etc, but I don't see what difference it makes when you are standing on the mound.

As others probably said better, I don't think Perkins uses most of these stats on the mound.  Reviewing PitchFX results is referenced in the article too, and I imagine that has more in-game practicality.

[Hosken Bombo Disco]

I can only hope that Jack Goin is force feeding some of this to the coaching staff.

Or better yet, directly to the players themselves.

[Otto von Ballpark]

It's just a shame that most of his teammates don't share his enthusiasm for the subject. This is a pretty decent primer on some of the more popular advanced metrics and how Perkins uses them to his advantage.

Emphasis mine -- I think this is why drjim was confused, thinking Perkins was employing these on the mound.  Most of the article deals with results like FIP, BABIP, LD%, etc.  Doesn't really give Perk an advantage, just some context when reviewing his past performance.  Correia could have equal awareness of these stats and still be incapable of pitching any better.

 

The PitchFX stuff, which is mentioned but not detailed in the article, is much more important, of course.  But I would hope reviewing PitchFX and applying lessons learned from it would be standard operating procedure at the big league level (if not minor league level too, in preparation for doing it at the big league level).  Is that's not the case here, then shame on the Twins.

[Otto von Ballpark]

I wonder if the Twins have their own means of searching, reviewing, dissecting Pitch FX and related data?  Perkins really shouldn't have to go to brooksbaseball.net after each appearance.  And if that's the only way to get that data in the Twins clubhouse, I would imagine that Gardy and Rick Anderson and most of the other players don't see it (which could be what Brock is alluding to here).

 

Now that I think of it, working with large data sets was a primary requirement for the recent front office job posting, so maybe that is finally in the works.  If that's the case, they've got to be behind the curve on this, though, and I'm not sure how much hiring a Visual Studio proficient programmer is going to help them...

[drjim]

Emphasis mine -- I think this is why drjim was confused, thinking Perkins was employing these on the mound.  Most of the article deals with results like FIP, BABIP, LD%, etc.  Doesn't really give Perk an advantage, just some context when reviewing his past performance.  Correia could have equal awareness of these stats and still be incapable of pitching any better.

 

The PitchFX stuff, which is mentioned but not detailed in the article, is much more important, of course.  But I would hope reviewing PitchFX and applying lessons learned from it would be standard operating procedure at the big league level (if not minor league level too, in preparation for doing it at the big league level).  Is that's not the case here, then shame on the Twins.

 

I'm not confused as much as curious on how applicable deep knowledge of stats are to players and their performance. And to be clear I am not against these stats in any way, they are really good to know for evaluation. I do think the overall knowledge of stats for the athletes specifically don't do much other than make tangible what is probably pretty intuitive for an athlete at this level. Though I suppose for more reflective types (like Perkins) it does provide positive reinforcement. I also think there can be value for confidence reasons if a specific player is not achieving results in traditional stats even if they feel they are pitching well (i.e. they are underperforming the underlying metrics).

 

I see pitch f/x as more a scouting tool than a statistical tool (perhaps that is unfair) that should be used by coaches/front office to improve performance and produce scouting reports on opponents to develop game plans. Perhaps I am splitting hairs.

[Mike Sixel]

do you all not work in companies with goals, objectives, and process control to understand how you are doing as a company (and as an individual contributing to the company)? Huh.

 

Also, it is odd, as someone said, that Perkins has to go to Fangraphs, and not "super secret Twins site" to get data......

[drjim]

Process control, like any other effort. He is doing statistical process control to understand the outcomes he is producing, to understand if he needs to change what/how he is doing it. Every successful business does this......

 

This makes sense, but I still see it more under the role of coaching or front office.

 

I asked my question in the context of the comment that "it's a shame his teammates don't show the same enthusiasm". I'm not certain it is that critical for a player to concern themselves with this. I'm skeptical it has much to do with Perkins being a better pitcher, I more credit his 95 fastball and filthy slider.

[drjim]

do you all not work in companies with goals, objectives, and process control to understand how you are doing as a company (and as an individual contributing to the company)? Huh.

 

Also, it is odd, as someone said, that Perkins has to go to Fangraphs, and not "super secret Twins site" to get data......

 

See my previous response. I am a little concerned that it is not more in house, makes me a little concerned about the actual scouting reports that are prepared for upcoming opponents.

 

In the first part - the goal is get outs and not allow runs, and the process is by limited walks and home runs, while maximizing ks and ground balls (to an extent). Again, I think it makes tangible what is intuitive, so it is definitely interesting and important for analyzing players for contracts, roster spots, etc., but I'm still curious what it actually means for specific players.

[Mike Sixel]

Oh, talent is a much bigger input to success than process control. I bet I could do SPC better than perkins (or about the same), but I have no chance to pitch.

 

Having been in communications and organizational change management and strategic management at various times, it is clear from my experience that employees benefit from having this kind of knowledge. We repeatedly find that letting employees know how their success is judged, how we measured their process performance, how the overall company's success was measured, greatly increased the quality, speed, and (decreased) the cost of outcomes produced.

 

For a baseball player, they have often looked at RBI or BA to judge how they were doing (in the old days, and still today). Pitchers (just ask jack morris) looked at wins. 

 

Perkins, he's looking at other outcome measures.

 

But, your overall point remains. Talent>process.

 

But, as Tony Gwynn reminded us, process is very important also.

[tobi0040]

Can Gardy throw 97 and close out games?  I am hoping for a swap

[Otto von Ballpark]

I'm not confused as much as curious on how applicable deep knowledge of stats are to players and their performance. And to be clear I am not against these stats in any way, they are really good to know for evaluation. I do think the overall knowledge of stats for the athletes specifically don't do much other than make tangible what is probably pretty intuitive for an athlete at this level. Though I suppose for more reflective types (like Perkins) it does provide positive reinforcement. I also think there can be value for confidence reasons if a specific player is not achieving results in traditional stats even if they feel they are pitching well (i.e. they are underperforming the underlying metrics).

Obviously these kind of stats don't matter much for an athlete's work within the actual competition.  And I don't think anyone here is saying that they do.

 

I think they appreciate that, when reviewing his own performances, Perkins uses stats that best reflect his actual contribution.  Dude's not hung up on saves or ERA (or pitcher wins for starters, or AVG/RBI for hitters), and is probably less likely to "rest on his laurels" by meeting a mostly meaningless benchmark.  (Although it's probably easiest to ignore "traditional" stats as a relief pitcher -- heck, if I played the game, I'd probably like to boost my BA from .299 to .300 on the last day of the season, or get my 100th RBI rather than take a walk, or gun for my 20th win even if I was supposed to rest for the playoffs or something. The influence of the backs of baseball cards has been strong!)

 

And, of course, prepping and game-planning with the aid of Pitch F/X should be mandatory, and it's good to hear someone on the Twins uses it.

[tobi0040]

Oh, talent is a much bigger input to success than process control. I bet I could do SPC better than perkins (or about the same), but I have no chance to pitch.

 

Having been in communications and organizational change management and strategic management at various times, it is clear from my experience that employees benefit from having this kind of knowledge. We repeatedly find that letting employees know how their success is judged, how we measured their process performance, how the overall company's success was measured, greatly increased the quality, speed, and (decreased) the cost of outcomes produced.

 

For a baseball player, they have often looked at RBI or BA to judge how they were doing (in the old days, and still today). Pitchers (just ask jack morris) looked at wins. 

 

Perkins, he's looking at other outcome measures.

 

But, your overall point remains. Talent>process.

 

But, as Tony Gwynn reminded us, process is very important also.

 

Peyton Manning and Ted Williams also come to mind.

[drivlikejehu]

Baseball players make adjustments to their game all the time. Accurate self-evaluation is absolutely critical... coaches certainly play a role, but these are professionals that have ultimate responsibility for their careers.

 

For instance, Perkins understands that having a high BABIP allowed probably doesn't mean something is wrong. He discussed that very issue early this season. Another pitcher might try to make a change that isn't necessary and might be counter-productive.

 

MLB is full of players with talent. The challenge is to get the most out of it.

[Hosken Bombo Disco]

I'd be really surprised if Perkins doesn't sift out helpful data on opposing batters and then brainstorm with Suzuki and the other guys. Maybe that's what he brings to the mound with him. The article didn't get into that -- and doesn't need to.

[adjacent]

See my previous response. I am a little concerned that it is not more in house, makes me a little concerned about the actual scouting reports that are prepared for upcoming opponents.

 

In the first part - the goal is get outs and not allow runs, and the process is by limited walks and home runs, while maximizing ks and ground balls (to an extent). Again, I think it makes tangible what is intuitive, so it is definitely interesting and important for analyzing players for contracts, roster spots, etc., but I'm still curious what it actually means for specific players.

I guess walks, home runs are part of the results too. The process would be subtle changes in the delivery, location, release point,etc. Of course Perkins is not going to talk specifically about those, because he does not want to give everything away, but when you compare those changes with a quantification of the results, that gives you a better idea of what it works, than just wether "it feels good".

[drjim]

Oh, talent is a much bigger input to success than process control. I bet I could do SPC better than perkins (or about the same), but I have no chance to pitch.

 

Having been in communications and organizational change management and strategic management at various times, it is clear from my experience that employees benefit from having this kind of knowledge. We repeatedly find that letting employees know how their success is judged, how we measured their process performance, how the overall company's success was measured, greatly increased the quality, speed, and (decreased) the cost of outcomes produced.

 

For a baseball player, they have often looked at RBI or BA to judge how they were doing (in the old days, and still today). Pitchers (just ask jack morris) looked at wins. 

 

Perkins, he's looking at other outcome measures.

 

But, your overall point remains. Talent>process.

 

But, as Tony Gwynn reminded us, process is very important also.

 

But do these stats actually help a player with process?

 

When I think of pitching, I would say:

Desired Outcomes = maximize outs, minimize runs

Desired Goals = maximize strikeouts, ground balls, minimize walks, home runs

Process = Preparation to pitch (work outs, bullpens, etc), scouting of opponent/game plan (scouting reports, pitch f/x data, understanding of location, game situations, etc), proper utilization (pitching roles)

 

I'm just thinking of how this specific knowledge fits in. I can see for confidence reasons and I suppose that specific personality types would eat it up more than others, but I hardly see it as necessary for most players.

[drjim]

I guess walks, home runs are part of the results too. The process would be subtle changes in the delivery, location, release point,etc. Of course Perkins is not going to talk specifically about those, because he does not want to give everything away, but when you compare those changes with a quantification of the results, that gives you a better idea of what it works, than just wether "it feels good".

 

I certainly don't dispute this, and like you said, he doesn't talk about it so it is impossible for us to know exactly what is happening in pre game preparation.

 

My question from this article is still - does a player having knowledge of advanced stats result in him performing better? I'm still skeptical.

[Monkeypaws]

I sense a future pitching coach in the Twins' organization....

[Oxtung]

I think that statistics should be used as a guide to what's wrong, or right, in what you're doing.  Once you've identified the problem you can use other tools, like video analysis or pitch f/x, to determine what specifically you need to change.  However the statistics are the smoking gun that something needs to change; or doesn't depending on what they say.

 

I actually think Perkins shows you how he uses it in this article.  First, look at FIP.  His is at 1.82 which is significantly below his ERA.  So why is that?  Is that just random fluctuation, is it his defense, is it something he needs to change?  Dive deeper, we see his babip is at .330.  Why is that elevated?  Look at his batted ball profile and we see that his LD% is slightly raised from career norms.  He's getting squared up more often than he has in the past.  Now he can look at more specific stats to determine when exactly that is happening.  What count are batters teeing off on?  Now you can go look at the pitch data from that specific count to determine which pitches are getting hammered.  Now you know specifically which pitch is struggling.  Now you can go to the video analysis, combined with the pitch f/x data, to determine what exactly you need to tweak to improve. 

 

I think that the statistics are the smoking gun that show something needs to change, or not.  After that they lead to what needs to be fixed.  They can't actually do the fixing themselves but they can tell you after the fact if your adjustments were successful.

[Parker Hageman]

I was able to chat with Perkins about his thoughts on stats prior to the home opener last year. He has lots of interesting stuff to say and questions about them as well. He also provided what facilitated his dive into the numbers:

 

“I didn’t get into any advanced stats until 2010 when I was in Triple-A and struggling,” said Perkins. “It was a combination of [Ryan] Vogelsong and when Brandon McCarthy came back and I started to hear about those guys who were injured or unsuccessful and then they adjusted their pitch selections and focused doing different things. That got me thinking, maybe it’s something I should do.”

 

 

You can read the rest here.

[TheLeviathan]

I certainly don't dispute this, and like you said, he doesn't talk about it so it is impossible for us to know exactly what is happening in pre game preparation.

 

My question from this article is still - does a player having knowledge of advanced stats result in him performing better? I'm still skeptical.

 

I'm sorry, but are you really leaving anyone the ability to answer this question or is this just an exercise in stubborn skepticism?

[JB_Iowa]

I was able to chat with Perkins about his thoughts on stats prior to the home opener last year. He has lots of interesting stuff to say and questions about them as well. He also provided what facilitated his dive into the numbers:

 

 

You can read the rest here.

 

Thank you Parker.  For anyone who hasn't read it, Parker wrote a great piece.